Relative Pose Estimation through Affine Corrections of Monocular Depth Priors
Yu, Yifan ; Liu, Shaohui ; Pautrat, Rémi ; Pollefeys, Marc and Larsson, Viktor LU
(2025)
p.16706-16716
- Abstract
- Monocular depth estimation (MDE) models have undergone significant advancements over recent years. Many MDE models aim to predict affine-invariant relative depth from monocular images, while recent developments in large-scale training and vision foundation models enable reasonable estimation of metric (absolute) depth. However, effectively leveraging these predictions for geometric vision tasks, in particular relative pose estimation, remains relatively under explored. While depths provide rich constraints for cross-view image alignment, the intrinsic noise and ambiguity from the monocular depth priors present practical challenges to improving upon classic keypoint-based solutions. In this paper, we develop three solvers for relative pose... (More)
- Monocular depth estimation (MDE) models have undergone significant advancements over recent years. Many MDE models aim to predict affine-invariant relative depth from monocular images, while recent developments in large-scale training and vision foundation models enable reasonable estimation of metric (absolute) depth. However, effectively leveraging these predictions for geometric vision tasks, in particular relative pose estimation, remains relatively under explored. While depths provide rich constraints for cross-view image alignment, the intrinsic noise and ambiguity from the monocular depth priors present practical challenges to improving upon classic keypoint-based solutions. In this paper, we develop three solvers for relative pose estimation that explicitly account for independent affine (scale and shift) ambiguities, covering both calibrated and uncalibrated conditions. We further propose a hybrid estimation pipeline that combines our proposed solvers with classic point-based solvers and epipolar constraints. We find that the affine correction modeling is beneficial to not only the relative depth priors but also, surprisingly, the "metric" ones. Results across multiple datasets demonstrate large improvements of our approach over classic keypoint-based baselines and PnP-based solutions, under both calibrated and uncalibrated setups. We also show that our method improves consistently with different feature matchers and MDE models, and can further benefit from very recent advances on both modules. Code is available at https://github.com/MarkYu98/madpose. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/32f0f5c0-4d66-4d4a-bd27-4e90ea514d5b
- author
- Yu, Yifan
; Liu, Shaohui
; Pautrat, Rémi
; Pollefeys, Marc
and Larsson, Viktor
LU
- organization
- publishing date
- 2025
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- pages
- 11 pages
- publisher
- IEEE
- ISBN
- 979-8-3315-4364-8
- DOI
- 10.1109/CVPR52734.2025.01557
- language
- English
- LU publication?
- yes
- id
- 32f0f5c0-4d66-4d4a-bd27-4e90ea514d5b
- date added to LUP
- 2026-04-02 13:58:08
- date last changed
- 2026-04-13 12:27:36
@inproceedings{32f0f5c0-4d66-4d4a-bd27-4e90ea514d5b,
abstract = {{Monocular depth estimation (MDE) models have undergone significant advancements over recent years. Many MDE models aim to predict affine-invariant relative depth from monocular images, while recent developments in large-scale training and vision foundation models enable reasonable estimation of metric (absolute) depth. However, effectively leveraging these predictions for geometric vision tasks, in particular relative pose estimation, remains relatively under explored. While depths provide rich constraints for cross-view image alignment, the intrinsic noise and ambiguity from the monocular depth priors present practical challenges to improving upon classic keypoint-based solutions. In this paper, we develop three solvers for relative pose estimation that explicitly account for independent affine (scale and shift) ambiguities, covering both calibrated and uncalibrated conditions. We further propose a hybrid estimation pipeline that combines our proposed solvers with classic point-based solvers and epipolar constraints. We find that the affine correction modeling is beneficial to not only the relative depth priors but also, surprisingly, the "metric" ones. Results across multiple datasets demonstrate large improvements of our approach over classic keypoint-based baselines and PnP-based solutions, under both calibrated and uncalibrated setups. We also show that our method improves consistently with different feature matchers and MDE models, and can further benefit from very recent advances on both modules. Code is available at https://github.com/MarkYu98/madpose.}},
author = {{Yu, Yifan and Liu, Shaohui and Pautrat, Rémi and Pollefeys, Marc and Larsson, Viktor}},
booktitle = {{2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}},
isbn = {{979-8-3315-4364-8}},
language = {{eng}},
pages = {{16706--16716}},
publisher = {{IEEE}},
title = {{Relative Pose Estimation through Affine Corrections of Monocular Depth Priors}},
url = {{http://dx.doi.org/10.1109/CVPR52734.2025.01557}},
doi = {{10.1109/CVPR52734.2025.01557}},
year = {{2025}},
}